8:u 



rOXTRU^.UTIOXS FKOM THE XATTOXAL lLKKr>ARIU-A[. 



for so assortiiii,^ ohjt^cts of uaiiire into small groups as to rt^icU'r easy tlu^ir availability 

 and idontififalion. Xor cdn at ihc prosont tinii* tliat otlu^r ])urpos(* of classilication, 

 the indiralion of relationships, l>o prrssod vory strongly. Rut unh^ss all th<* usual 

 conc(^ption> of the word 'gonuy' aro to Ix^ ignored. tlnMlistinctions drawn l)y\\'illiani 

 Smith must ht^ t'onsitU'n'd as loo uniniportant anil ospoeially loo unslabh' to admii (if 



4 



adoptiijn. 



By fai- the most [winslaking aiKl thorough analysis of Xavicula is thai ui (devc in 

 his Navi<'uloid Diatoms/^ I regrc^t I can not ft)llo\v his <livisi(>ns of this g(Muis. It 

 would \)c asid<' from the purpose of this n^port to dis(Miss in aiujile detail the merits of 

 this (^xtensive work. Tt must, th(M-efore^ sufliee to say thnt a ran^ful reading of the 

 ehapt(M' "On the value of the eharaet(^risti(^s," on pag(^ 1 of that w^ork, wdll, in eonriee- 

 tion witli th<^ divisions of Navicula sul)se(ja(uitly made in the work, reveal the fact 

 that lu^arly all the eharacteristies there looked upon as inadequate are subsetpu^ntly 

 employed lvs l)as(^s for the n( w genera created. I th(^refore follow in gpueral the 

 majority of authors writ ing sine(^ rh^v<''s work aj)peared, in eonsidering these distine- 

 tions of vahie for subg<^nerie grouping, though not of generic worth. P^speeially has 

 Van Il^nu-ek ^-' grouped tlie geniis Navieula in a way sat isfaetory to the writer. Exet^j)- 

 tion must, however, he taken to Ids r(*cognition of tin* g(Mnis Staunmeis Khn-nh. 

 as a g(^nus. As Van Ileurrk iIhto says, the true Stauroneis forms differ in 

 no respect from Xavicula, except ''by the central noduh^ being transversely dilated 

 into a stauros." This has always appeared to me to be a very trivial ground. The 

 stauros is a most conunon accident of iuany species in several other genera, as A(h- 

 narUlu^s li(»ry, in .1. roarclafa (13rel>/j Crun.;^ Pleurosigma W. Smith, in P. atiiadann 

 Temp. & lirun auil P. sfaurophonim firun.;^ and CocconrMS Ebrenl>., in C. formosn 

 Hrun/' Xor is lliis any more or less the ^'dilation of the central nodule- in the on(^ 

 case than in th(^ o1h(*r. In short, the separation of stauros-bearing forms into a genus 

 is an impossibility, oitlier inside the gcmus Xavicula or outside^ of it. 



I look upon the following genera as having fairly good claim to s(^parate standing: 

 Maslogloia Thwaiti's, for reasons stated under that gt^nus; Diclyoneis Cleve. on 

 account of its peculiar internal plates and (^sp(^cially its loculato border,/ though it 

 mav possiblv be necessarv to unite it with Masto^'htia; llouxia Brun & Ilerib.; {/ also 

 the following six genera, which, allhough sc^jiaratt^ from Navicula, should be united 

 under the sanu^ generic name: FrustuliaC Ag., in l>art, Berkeleya (Gr(^v.) Van Tleur.^ 

 Reii'holtia Van Ileur., Amphipleura Kiitz.^ Rr<'l)issonia (Jrun., Vanhenrckia Hreb.^ 

 for all whicli see under Frustulia in this repr>rt ; also tluMr figures and descriptions 

 bv Van Ileurck. ^ 



To the foreg{)ing may be added Chive's genus Vistula:? for although it at present 

 in( lu(h^s only the single species C. Inrrnziana (Grun.) rie\'e, (his is perf(H'tly constant 

 and strikingly unlike other navicuh)id diatoms. On the wliole the best conception of 

 Na\ icula and its allies is that of Van Ileurck. ./ 



Navicula aestiva Donk. Trans. Micr. Soc, T.ontl. u. s. 6: .12. ph S,f, IS. 1858; P>rit. 

 Diat.G. pL l.f. .?. 1871^73. Schmidt, Atlas y>/. ;./. ,^J() 11 {uwn-.uncd). pi. S. f, ?6 

 (type),/ Jl (unnamed). 1870. Ral>h. Fb Eur. Alg. 1: ISl. ISGI. DeToni, Svlb 



(r Sv. V<4. Akad. Ilandb 26-: 181)4; 27^^: ISDo. 



b V,n Heur. Treat. Diat. 189(1. 



^- Van Ileur. Synop. pLJdJ. IS. ISSl.' 



fi PiTag. be Diatomist<' 1*-"': pJ.S.f.-i'i li. ISiM. 

 *• Schmi<lt, Alhis pJ. }!Ki,f,4G, |S1)-1, 



/('f. Sv. Vet^ Akad. Haiulb 26': 1:^4. bsu4: and Van Heur. Treat- Diat. 157. JHOH. 



£7 Ilerib. Diat. Auverg. Io(>. 18*K5. 



ft Van Hour. Treat. Dial. 2:V.), 1^2-245. ISDH. 



1 Sv. Vel. Akad. TIandb 26 : 124. 1S94. 



jX-jn Ileur. Treat. Dial. I(i2-2S7. 189(1. 



